Raspberry Pi delivers great multi-room audio

Raspberry Pi with IQaudIO DigiAMP+

Raspberry Pi can be at the centre of great-sounding connected home audio

There was a time when multi-room audio – streaming music to multiple rooms in the house with control from anywhere – was only for the well-off, or the extraordinarily ingenious.

Even mid-range systems were £300 per room.

Raspberry Pi has changed all that, and now the only qualification for such a system is the ability to follow step-by-step instructions – and maybe £100 per room for good sound through existing loudspeakers, and under ~£30 if you have an old set of computer speakers lying around and are really not fussed about quality.

Even though they are credit card sized, some Raspberry Pi hi-fi add-ons include Class-D power amplifiers delivering tens of watts. And it is not only makers that turn to Raspberry Pi, there are professional bespoke systems out there constructed around the little educational computer.

Hardware

Raspberry Pi has had PWM-based on-board audio output capability since it was first released. It is great for the computer’s educational remit, perfect for beeps and tones, and will play music through active PC loudspeakers, but will never set the audio world on fire.

Almost as soon as the audio output limitations were noticed, members of the open-source community looked for ways – both hardware and software – to add external music-quality DACs.

Initially, these had to be USB DACs because that was the only available interface, but the first major revision of Raspberry Pi’s PCB bought the host chip’s existing I2S audio bus out to an 8pin header (P5), and these connections moved to the main 40pin I/O when it replaced the original 26pin I/O connector when the ‘plus’ and subsequent ‘2’ Raspberry Pis emerged.

Invented by Philips, I2S is a three-signal (clock, data and left/right) serial interface dedicated to digital stereo. It is agnostic with respect to sample rate or sample word length and was intended as a way to send data to dumb two-channel DACs. Provided the word length (usually 16bit) matches, the DAC will play at whatever sample rate data is sent.

There are also smart audio chips that take I2S as input and emit analogue audio. These generally need a separate bus connection to set them in the right mode to function like a dumb DAC.

Several companies have taken DACs and built them into audio cards that sit on top of Raspberry Pi and connect to the I2S bus.

Amongst them are cards from Element14/Cirrus/Wolfson, Modul 9 (HiFiBerry) of Switzerland and IQaudIO.

Gordon Garrity is founder of IQaudIO, which both designs and makes its Hi-Fi audio boards in the UK.

“People can plug in USB sound card, and there are good ones for under £200, but the [Raspberry Pi] USB chip shares a bus with Ethernet and an overloaded bus was a problem with original Pi for HD audio,” he said. “This should be less of a problem with the quad core [Raspberry Pi 2].”

IQaudIO Pi-DAC+ for Raspberry Pi

Pi-DAC+ from IQaudIO is a I2S-connected Raspberry Pi HAT with line and headphone outputs

Another issue, according to Garrity, is that a lot of USB sound cards cannot convert faster than 96kbit/s as this is the maximum for Windows drivers. Expensive USB DACs have 192kbit/s chipsets and by-pass the Windows drivers. Linux and Macs will handle these 192kbit/s files and DACs.

And they protrude outside Pi’s svelte credit card profile.

All of which is why his company uses the I2S connection.

Getting clean sound when bolted to a computer can be tricky – Raspberry Pi is quite noisy, for example.

“We try to lay boards out to get the best audio we can,” said Garrity. “The PCB needs a good ground plane, and we have a lot more components than some audio cards to stopping us from being susceptible. There are some on the headphone output, and some stopping conducted noise.

Grounding is another issue that has had careful attention to avoid loops – made more complex from the Pi ‘+’ versions onwards because the computer can be powered directly or powered from its add-on board.

And however careful you are trying to be, some things will still show up if you have an audio analyser to hand.

“RaspPi is the I2S master and its clock is not brilliant. There is a bit of jitter, and you are not going to get jitter-free with a sub-£100 device,” said Garrity. “The Raspberry Pi market cannot sustain a £150 selling price and the target price has to be as close to the Raspberry Pi price as possible.”

Farnell (Cirrus/Wolfson), IQaudIO and HiFiBerry sell DAC add-on boards, with IQaudIO and HiFiBerry also selling single boards that combine a DAC and Class-D amplifier delivering over 20W/channel stereo. Some of these have headphone outputs as well.

Farnell’s offerings include a 2+2W power amp, and differ from the others in having high-quality audio input ADCs, mems microphones and sophisticated mixers, plus SPDIF digital interfaces.

Low-level software

None of these audio products can be usefully attached to a Raspberry Pi without some modification to the Pi’s software. Even a dumb DAC can need a minimal driver, just to say it has been attached. More complex DAC hardware may need data loaded into internal registers to set it into the appropriate mode – I2C bus is a favourite amongst chip makers for this, and Raspberry Pi’s Broadcom chip has an I2C bus, brought out onto the 26/40 way connector.

The Cirrus/Wolfson boards are supplied with scripts to run which set them into various modes via I2C.  IQaudio hardware also uses I2C to configure the DACs, with the necessary I2C device drivers built into the Raspberry Pi Linux release (Raspbian).

Deep within the Linux Kernel is an audio subsystem called ALSA (Advanced Linux Sound Architecture which can receive and send on multiple channels of digital music and mix between them.

We have German student Florian Meier to thank for working out how to enable the I2S hardware within the Broadcom chip at the heart of Raspberry Pi, and get it to work with ALSA. His code has been in the Linux main line kernel since version 3.14.

Raspberry Pi with IQaudIO DigiAMP+

20+20W Class-D stereo is available in HATs from more that one manufacturer. This one is Pi-DigiAMP+ from IQaudIO

There is also something called ASoC (ALSA System on a Chip), which is a subsystem that provides a framework for audio in embedded systems running Linux.

Other people have created device drivers for specific chips. According to Garrity there are some 100 chip drivers shipped along with Linux, some of them supplied by Edinburgh-based Wolfson, now absorbed into Cirrus logic.

IQaudIO uses a Texas Instruments audio DAC/codec in its Pi-DAC+ product. “We created a driver for the TI PCM5122. I wrote a proof of concept driver and Mark Brown [ex Wolfson] and Florian generously created a full device driver [PCM5122x.c],” said Garrity. “As it’s open source, others have continue to add new features and improve the quality.” Pi-DAC+ has line-level and headphone outputs.

IQaudIO’s 20W+20W DAC/amplifier called Pi-DigiAMP+ uses the same driver because it is based around TI’s TAS5756m which shares a register configuration with the PCM5122 – it is functionally equivalent to a 5122 followed by a TPA3118 Class-D amplifier. As it happens, IQaudIO also uses the 3118 in a product called Pi-AMP+ which fits over the Pi-DAC+ to add loudspeaker driving capability without losing line out and headphone support.

“The Raspberry Pi went through changes in January. We wanted to make sure our device drivers were in the standard kernel, then you only need to configure the through a text file,” said Garrity – whose boards are the ones that can be found built into Abacus Electronic’s bespoke audio products in Germany.

Raspbian

Having a Pi, and audio board, a kernel and drivers means you have a computer capable of outputting sound, but the computer has no idea what file to play and where to get it from.

If you are just playing music while using a Pi as a desktop computer, Raspbian with suitable drivers will do the job.

In general, most drivers for the various audio add-on boards for Raspberry Pi are supplied as part of the Raspbian operating system, but not for all.

This gives you two choices: add the drivers to Raspbian or download a pre-modified SD Card image from the board maker – in this last case, you have to keep your fingers crossed that the board maker is updating their image with each new Raspbian release.

GitHub is your friend for obtaining compile-it-yourself drivers. And some of the manufacturers have step-by-step instructions to walk you through the compilation process.

Be warned, executing a sudo apt-get update is likely to remove all the customisation.

The Raspberry Pi Foundation included an EEPROM in the specification of the latest generation of add-on boards (HATs – hardware attached on top), that will one day mean updates will automatically retrieve packages suitable for attached hardware.

“We have the EEPROM on our boards, but it is blank at moment. The Foundation pretty much has it all worked out now. It is not something that is blocking anything.”

Special music playing software

Up to this point, the computer is desktop PC-like, with hardware set up through the command line.

Raspberry Pi comes into its own as a music player when pre-compiled custom Linux media players are installed – see the list.

These have been created by people from the open-source community and can include all the features expected from an advance sound system – playback of local and network music files, streaming from on-line services like Spotify, Internet radio, remote control from phones and tablets, and ‘headless’ playback where the Pi is a network black box with loudspeakers but no local keyboard or display. Some software solutions even emulate other manufacturer’s consumer electronic solutions such as Apple Airplay and Logitech’s squeezebox. Take a look at the list at the end of this article.

Frequently in these cases, setting up for the appropriate DAC hardware is just a case of selecting it on a drop-down menu.

Hardware makers and software creators know each other.

“A bunch of people around the world are writing nice interfaces for audio play-back,” said Garrity. “The software providers are not monitising – it’s free to download. They are doing something to help me, so I provide them with hardware free-of-charge, and we support the guys who take donations.”

 

Raspberry Pi and Wolfson audio cardElectronics Weekly just happened to have a Raspberry Pi Model B (P5 connector version) and a compatible Wolfson Audio card from element14 in its sample cupboard.After downloading suitably-modified Raspbian from element 14’s website and programming an SD Card, it only took a matter of minutes to get it working – you have to run a supplied configuration script on first use.Included in the download were some HD audio files, and these sounded great through a fair amplifier and a nice pair of speakers – clear, sparkling, no hum, no hiss. On-line reports suggest IQaudIO and HiFiBerry products also sound great, and that people have underestimated just how good Raspberry Pi can sound with the right add-on.

 

Music on Raspberry PiSome music player software for Raspberry Pi

  • PiMusicBox
  • Volumio
  • RuneAudio
  • Max2Play
  • MoodeAudio
  • Squeezelite
  • Archphile
  • Openelec
  • Kodi
  • Shairport-Sync (Airplay renderer)

steve bush

Toshiba adds high luminous flux GaN-on-Si LEDs

Toshiba has added four GaN-on-Si LEDS to its TL1L4 series which provide a high luminous flux and are suited to applications ranging from street and stadium lighting to LED light bulbs and down lighting for use in the home. imageAll offer a colour rendering index of Ra70 min. and a typical forward voltage of 2.8V.

Like the existing members of the TL1L4 series, GaN-on-Si technology has been used to create LEDs optimised for both output and energy efficiency.

However, the new 4A5B type improves upon these existing products by providing a luminous flux of 140lm (min at 85°C) compared to 130lm (min).

The products make it possible to meet the market demands for improved lighting fixture efficacy by achieving efficacies of over 110lm/W in a finished luminaire – with Toshiba assuming 90% driver efficiency, 90% optical efficiency, and junctions at 85°C.

Measuring 3.5 x 3.5mm, with a lens-top, the LEDs offer differing correlated colour temperatures (CCTs), with the TL1L4-DW0 providing 6,500K, the TL1L4-NT0 providing 5,700K, the TL1L4-NW0 5,000K and the TL1L4-WH0 4,000K.

See alsoImec extends GaN-on-Si R&D programme

More Toshiba stories on Electronics Weekly »

 

david manners

How will consumers set up IoT security?

Soton IoT security

Research IoT node that can receive security parameters from a phone is multiple ways

University of Southampton researchers have studied ways to ease the arrangement of IoT devices in the home.

The issue is that to join a wireless network, security parameters have to be accurately entered into the IoT device, which may have almost no user interface.

“IoT objects can be attacked and possibly hijacked, putting our privacy, data and safety in question. We believe that not enough attention has been placed on how to make the IoT easy to use and to configure,” said Dr Enrico Costanza of the Agents, Interaction, Complexity Group at the University.”

Versions of four basic techniques were tried, all using a touch screen phone as an interface to enter a password, which the phone then send to the IoT device:

  • Holding the phone’s screen against the IoT device and transferring the parameters to a photo cell by flashing the phone screen.
  • Using a USB cable between phone and IoT device
  • Using a 3.5mm jack cable between the phones headphone socket and the IoT device
  • Using the phone to create a temporary ad-hoc Wi-Fi network which the IoT joins automatically.

30 untrained subjects were asked to try the techniques, and it was “found that two of the techniques were noticeably more usable than the others – the audio cable and the Wi-Fi-only interactions”, said the University.

“We believe that our results can help designers and researchers make IoT devices, and especially their configuration, more usable and therefore secure. We hope that our results will motivate others in researching this topic,” said Costanza.

The results are presented at the ACM Ubicomp 2015 conference in Japan this week.

All four IoT password transfer techniques are shown in this video.

 

steve bush

ST chip could turn TV services boxes into Wi-Fi hubs

Your TV services box could also soon become your Wi-Fi hub. STMicroelectronics has announced an HDTV set-top box system-on-chip  device with integrated Wi-Fi functionality.

The chip firm has used Wi-Fi 802.11 IP developed by Quantenna in its latest Cannes Wi-Fi (STiH390) SoC.

STiH390 - ST chip could turn TV services boxes into Wi-Fi hubs

STiH390 – ST chip could turn TV services boxes into Wi-Fi hubs

This is an 28nm FD-SOI chip with a multi-core ARM CPU capable of delivering 6K DMIPS and Mali 400 GPU for 3D graphics.

Quantenna’s baseband IP supports 4×4 802.11ac/11n wave2 Wi-Fi performance for video delivery in the home.

According to Philippe Notton, general manager of consumer products at STMicroelectronics, the company has validated 4×4 Wi-Fi 802.11ac on existing video-distribution.

Another feature is support for High Dynamic Range (HDR) content decode and display.

The company describes the chip as follows:

The device integrates leading ARM application processors architecture and GPU to provide thin client platforms, or interactive broadcast set topbox
(STB) platforms, supporting the latest middleware and software solutions.

The device’s integrated carrier-grade fullyoffloaded Wi-Fi MAC allows full HD video streaming throughout the home, making it the ideal device for Wi-Fi client boxes. The device supports full HD, high-efficiency video coding (HEVC) reducing memory bandwidth for video distribution.

Read more STMicroelectronics stories on Electronics Weekly »

 

Richard Wilson

Intepro launches auto-ranging power source

Intepro of Sandhurst, the power ATE specialist, has introduced an Auto-Ranging Programmable DC Power Source.

Model PSI 9000 provides up to 15kW in 3U rack-mount chassis. The auto-ranging feature maintains full rated power output at lower-voltage settings; units can be paralleled up to 150kW

A single 3U chassis houses up to 15kW of DC power and can be paralleled up to 150 kilowatts. Each chassis features a controller that allows the flexibility of separating into individual sources or paralleling for high-power applications.

Intepro launches auto-ranging power source

Intepro launches auto-ranging power source

A feature of the PSI 9000 Series is its auto-ranging output. Unlike DC sources that provide rated power only when maximum voltage is applied to the load, the auto-ranging output stage is capable of delivering three times higher output current at reduced voltages.

This feature is especially useful when testing products that require varied input voltages while maintaining regulated output power. With auto-ranging, a single-system provides a complete test solution compared to buying multiple sources to address low- and high-voltage/current requirements.

The PSI 9000 Series features an interactive touch panel menu navigation system that simplifies set up and storing to test profiles. The industry leading high-efficiency unit (up to 95.5%) includes an integrated true function generator. Complex test sequences can be loaded from and saved to a standard USB flash drive. The unit also includes simulations for PV (photovoltaic) arrays and fuel cells.

“The PSI 9000 Series is designed to simplify test set up and operation,” said Gerard Sheehan, Intepro’s CTO, “Front panel controls include an intuitive LCD display and control knobs for setting voltage, current, power or internal resistance simulation values.”

Additional features of the RoHS-compliant PSI 9000 Series include: a galvanically isolated analog interface for voltage, current and power programming and monitoring, digital plug-and-play modules supporting RS232, Ethernet, CANopen, Modbus TCP, Profibus, Devicenet and IEEE/GPIB, and SCPI command language support.

The PSI 9000 Series is offered in models with output power ratings of 0-3.3kW, 0-5kW, 0-10kW and 0-15kW in a single 3U chassis. The PSI 9000 3U family includes eight voltage models ranging to 1,500VDC.

david manners

Diamond Microwave adds to GaN PA range

Diamond Microwave of Leeds, the high performance microwave power amplifiers specialist, has announced a ten-fold increase in its power output capability with the addition of a 1kW X-band model to the company’s range of GaN-based pulsed solid-state power amplifiers (SSPAWimage).

The DMX1K001 is an ultra-compact SSPA operating over a 1200MHz bandwidth centred at 9.5GHz. The ultra-compact design measures only 244 x 134 x 50mm3 excluding heatsink and connectors, making it ideally suited for use as an alternative to a travelling wave tube amplifier (TWTA) in various radar applications.

“Like all our GaN SSPAs, these amplifiers are extremely compact, employing a chip-and-wire microwave design,” said Ian Davis, business development manager at Diamond Microwave. “This state-of-the art performance is combined with a power-to-volume ratio we believe to be among the highest in the industry for such products. Similar designs can be tailored to suit other frequency ranges in the 1-18GHz range.”

The Diamond Microwave amplifier range features designs that are flexible in layout and architecture, and are fully customisable to meet individual specifications for electrical, mechanical and environmental parameters. All of Diamond Microwave’s amplifiers are suited to demanding defence, aerospace and communications applications.

david manners

Quantenna sampling 10G Wave 3 WiFi for 10Gbps

Quantenna is sampling 10G Wave 3 Wi-Fi products built on Quantenna’s True 8×8 QSR10G Wi-Fi platform with multi-user MIMO (MU-MIMO) technology for home wireless access points and residential gateways.

Quantenna’s up-to-12-streams 10G Wave 3 can address both the service provider and retail market segments.

10G Wave 3 WiFi supports speeds up to 10Gbps attained by combining True 8×8 MIMO configuration for 5GHz networks with a 4×4 MIMO configuration on 2.4GHz networks for a combined 12-stream configuration.

Sam Heidari - Quantenna CEOQuantenna’s 10G Wave 3 platform delivers the maximum capacity within the minimum spectrum so that more streams can be simultaneously transmitted with improved range and reliability.

This means that 802.11 Wi-Fi access points or gateways using 10G Wave 3 can support more clients, each running more efficiently for such demanding applications as real-time video, OTT video, voice, IoT and other services.

Single-user MIMO (SU-MIMO), 1×1 and 2×2 client performance is also improved, ensuring that the millions of currently shipping mainstream smart phones, tablets and notebook PCs will also benefit from an improved Wi-Fi experience.

The QSR10G Wave 3 product family also supports a unique adaptive MIMO architecture. This enables 10G Wave 3 access points to maximize overall network performance, delivering the best possible capacity across all client devices. This is important as the mix of legacy and new client devices are changing, as end consumers adopt more devices with 802.11ac.

Key Wave 3 features and Benefits
• Integrated AP chipset for dual-band (5GHz and 2.4GHz), dual concurrent operation and management
• 160MHz channel support for 5GHz networks. When combined with 8×8 MIMO configuration, this offers 4x the capacity of 80MHz 4×4 MIMO networks
• Unique adaptive MIMO configuration
• Up to 1024 QAM modulation for 2.4GHz and 5GHz transmissions
• Supports rich set of interfaces to external hosts such as PCIe Gen3/Gen2, RXAUI, RGMII, and others

The QSR10G product family delivers 10G Wi-Fi with up to 8×8 MU-MIMO for smart access points, wireless cable/DSL gateways. Key features of the industry’s first 10G Wave 3 Wi-Fi product family include:

• QSR10GU
o Peak PHY rate of near 10Gbps
o Support for simultaneous 5GHz and 2.4GHz networks
o 12 stream operation
• QSR10GA
o Peak PHY rate of 9Gbps
o Support for simultaneous 5GHz and 2.4GHz networks
o 10 stream operation
• QSR10PA
o Peak PHY rate of 7Gbps
o Support for simultaneous 5GHz and 2.4GHz networks
o 8 stream operation
• QSR10G5
o Peak PHY rate of 8.6Gbps
o Support for 5GHz networks
o 8 stream operation

“In Wi-Fi access points more antennas are always better, but 10G Wave 3 is not just about 8×8 MU-MIMO and faster speeds. It’s about making better use of network and airtime efficiency to support the growing number of connected devices, services and applications,” said Dr. Sam Heidari, Quantenna CEO.

david manners

National Coding Week aims at adult learners

National Coding Week is intended open up coding to adults, potentially leading to careers in programming.

Taking place over September 21-27, there will be free-to-attend training sessions across the UK, where people will be taught the basics of coding alongside other beginners with little or no previous knowledge. There is a clickable event map on the National Coding Week website.

“The key aims of National Coding Week are to encourage adults of any age to learn an element of computer coding, to encourage digital experts to share their skills, and to collaborate, share, learn and have fun,” said event co-founder Richard Rolfe. “If I can learn to code at age 51 then anyone can.”

According to organisers, 52% of adults in the UK said they would feel ‘extremely or very daunted’ at the prospect of learning to code, while 41.3% believed that the older they are, the harder it is to pick up a skill like coding.

Boris Johnson is supporting the event for a second year.

“Coding is a language that is increasingly important for both young people and adults to understand, but it can be an intimidating prospect. National Coding Week will help to make whole generations of Brits more comfortable with coding, allowing them to embrace the business opportunities of the future,” he said.

Organisations already supporting this year’s event include: Decoded, Women Who Code, We Got Coders, Codeclan, Incubus London and International tech conference organisers Future Insights.

“The future is being written in lines of code.  Whether you’re 7, 17, 37 or 70 there has never been a better time to learn the languages of technology, the language of billions. Take the leap during National Coding Week,” said Decoded co-founder Kathryn Parsons. “Whether you sign up to a code school, learn on-line or buddy up with a coder in your local community, there are so many ways to learn. Take your first steps behind the screen, unleash your inner geek and see how fun and creative it can be.”

This year’s event is sponsored by JT Group Global.

steve bush

National Coding Week aims at adult learners

National Coding Week 2015National Coding Week is intended open up coding to adults, potentially leading to careers in programming.

Taking place over September 21-27, there will be free-to-attend training sessions across the UK, where people will be taught the basics of coding alongside other beginners with little or no previous knowledge. There is a clickable event map on the National Coding Week website.

“The key aims of National Coding Week are to encourage adults of any age to learn an element of computer coding, to encourage digital experts to share their skills, and to collaborate, share, learn and have fun,” said event co-founder Richard Rolfe. “If I can learn to code at age 51 then anyone can.”

According to organisers, 52% of adults in the UK said they would feel ‘extremely or very daunted’ at the prospect of learning to code, while 41.3% believed that the older they are, the harder it is to pick up a skill like coding.

Boris Johnson is supporting the event for a second year.

“Coding is a language that is increasingly important for both young people and adults to understand, but it can be an intimidating prospect. National Coding Week will help to make whole generations of Brits more comfortable with coding, allowing them to embrace the business opportunities of the future,” he said.

Organisations already supporting this year’s event include: Decoded, Women Who Code, We Got Coders, Codeclan, Incubus London and International tech conference organisers Future Insights.

“The future is being written in lines of code.  Whether you’re 7, 17, 37 or 70 there has never been a better time to learn the languages of technology, the language of billions. Take the leap during National Coding Week,” said Decoded co-founder Kathryn Parsons. “Whether you sign up to a code school, learn on-line or buddy up with a coder in your local community, there are so many ways to learn. Take your first steps behind the screen, unleash your inner geek and see how fun and creative it can be.”

This year’s event is sponsored by JT Group Global.

steve bush

ADI launches clock jitter attenuator for base-stations

Today ADI introduced a clock jitter attenuator designed to support the JESD204B serial interface standard for connecting high-speed data converters and FPGAs operating in base station designs.

The JESD204B interface was developed to address high-data rate system design needs, and the 3.2-GHz HMC7044 clock jitter attenuator contains functions that support and enhance the unique capabilities of that interface standard.

ADI HMC7044 block diagram

ADI HMC7044 block diagram

The HMC7044 delivers 50-fs jitter performance, which improves the signal-to-noise ratio and dynamic range of high-speed data converters, and the device provides 14 low-noise and configurable outputs that provide flexibility in interfacing with many different components.

The HMC7044 also offers a wide range of clock management and distribution features that make it possible for designers of base stations to build an entire clock design with a single device.

In base stations applications there are many serial JESD204B data converter channels that require their data frames to be aligned with an FPGA. The HMC7044 clock jitter attenuator simplifies JESD204B system design by generating source-synchronous and adjustable sample and frame alignment (SYSREF) clocks in a data converter system.

The device features two phase-locked loops (PLLs) and overlapping, on-chip, voltage-controlled oscillators (VCOs). The first PLL locks a low-noise, local voltage-controlled clock oscillator (VCXO) to a relative noisy reference, while the second PLL multiplies the VCXO signal up to the VCO frequency with exceptionally little added noise.

For cellular infrastructure JESD204B clock generation, wireless infrastructure, data converter clocking, microwave baseband cards and other high-speed communications applications, the architecture of the HMC7044 offers excellent frequency generation performance with low phase noise and integrated jitter.

HMC7044 Clock Jitter Attenuator Key Features

• JEDEC JESD204B support
• Ultra-low RMS jitter: 50 fs (12 KHz to 20 MHz, typical)
• Noise floor: -162 dBc/Hz at 245.76 MHz
• Low phase noise: < -142 dBc/Hz at 800 kHz to 983.04 MHz output frequency
• Up to 14 device differential device clocks from PLL2
• External VCO input supports up to 5 GHz
• On-board regulators

The device costs $12.75 in a 10-mm × 10-mm LFCSP package

david manners